{"id":"9318fcbd-004d-4861-8c8e-fec152ac4280","arxiv_id":"1908.03347","paper_version":1,"verdict":"CONDITIONAL","confidence":"LOW","novelty_score":7.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"For finite G=AB, the condition that <a,b> is soluble for all a in A and b in B is equivalent to the commutator [A,B] lying in the soluble radical of G.","lead":"For finite groups that are products of two subgroups, this paper shows that a local condition, every pair of elements from opposite subgroups generates a soluble group, is equivalent to a global condition, the commutator of the two subgroups being soluble. This extends Thompson's theorem and gives a local-to-global test for solubility in factorized finite groups.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Load-bearing concern: Theorem 1's almost-simple case analysis is explicitly incomplete; a single wrong independence claim in Tables 1–14 would invalidate the Main Theorem.","rationale":"The reader identified the same weakest assumption, and I agree. The paper makes an explicit self-limitation: Section 2 is an outline, and the omitted checks are exactly the load-bearing part. I considered whether a more specific internal gap exists in Section 3; the minimal-counterexample reduction (steps (i)–(vii)) is coherent, the use of external results [16,22,23,24,25,26,36] is appropriate, and the final order argument in (vii) is sound. No machine-checked proofs or code accompany the paper, so the classification-based tables are the only evidence for Theorem 1. This does not mean the theorem is false; it means the correctness of the Main Theorem is inherited from a collection of unverified finite and infinite family claims. Hence the appropriate disposition is to keep the reader's CONDITIONAL verdict: accept provided the table audit succeeds. If the audit shows a counterexample to a row, the verdict would move to REJECT; if the audit fully confirms the tables, it could move to ACCEPT.","tokens_in":24505,"tokens_out":18930,"duration_ms":190470,"concrete_test":"Run a computational audit of the finite exceptional rows and the two special cases flagged in Section 2: for each row of Tables 2, 4, 5–10, 12–14, and for U4(3) in §2.2.2(m) and PSp6(2) in §2.2.3(g), use GAP or Magma with the ATLAS of finite groups and the available maximal-subgroup/table-of-marks libraries to verify (i) the listed prime divides the corresponding intersection A∩N or B∩N via the stated maximal-subgroup orders, and (ii) no soluble subgroup of N has order divisible by the product of the two listed primes (e.g. by checking the possible subgroup orders or normalizer structure). For the infinite families, re-derive the independence claim for each 'handled exactly' row from Lemmas 6–9 of [2] on at least one representative parameter set. If even a single row fails, Theorem 1 and hence the Main Theorem are not established.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The Main Theorem is reduced in Section 3, step (iv), entirely to Theorem 1 for almost simple groups, and the proof of Theorem 1 is a case-by-case analysis over the Liebeck–Praeger–Saxl classification [34]. The paper itself states, immediately after Lemma 2, that 'usually detailed checking work and easy calculations are omitted.' Only the linear case a1) is shown in detail; most other entries in Tables 1–14 are approved with phrases such as 'handled exactly as a1)', 'easily checked', or 'by [13]'. For instance, the U4(3) case in §2.2.2(m) depends on a uniqueness claim about maximal soluble subgroups containing an order-7 element that is not derived in the text, and the PSp6(2) case in §2.2.3(g) rests on a subgroup-order computation that is merely asserted. Each row must be correct: the minimal-counterexample reduction forces an almost simple configuration, so if any row misidentifies the two primes or the solubility obstruction, a genuine almost simple counterexample would survive and the equivalence in the Main Theorem would fail. The argument is structurally plausible and I found no explicit contradiction, but as printed the proof is not independently checkable without redoing the classification data.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper proves a Thompson-type characterization for factorized finite groups. The Main Theorem states that for a finite group G=AB with subgroups A and B, the following are equivalent: (1) A and B are S-connected, i.e. every subgroup <a,b> with a in A and b in B is soluble; (2) for all distinct primes p and q, every p-element of A and every q-element of B generate a soluble subgroup; and (3) [A,B] is contained in the soluble radical G^S of G. The proof proceeds by a minimal-counterexample argument in Section 3, reducing the problem to almost simple groups; Section 2 then proves Theorem 1, which says that in an almost simple group N <= G = AB <= Aut(N) with AN=BN=G and with A,B satisfying condition (2), one must have A=G or B=G. Theorem 1 is established by a case analysis over the Liebeck–Praeger–Saxl classification of maximal factorizations, using Lemma 2 for p-divisibility of A∩N and B∩N, and using independence of primes with respect to the soluble graph. A byproduct, Theorem 2, records the existence of independent primes in such factorizations, with exceptions for PSp6(2), U4(3), and several small Lie-type groups.","tokens_in":24733,"tokens_out":16908,"duration_ms":178175,"significance":"If the main theorem is correct, it is a substantial contribution: it simultaneously generalizes Thompson's theorem on solubility of two-generated subgroups, Carocca's theorem on products of S-connected soluble subgroups, and the Guralnick–Kunyavski–Plotkin–Shalev criterion for membership in the soluble radical. The Section 3 reduction is elegant and mostly self-contained, and it makes clear why the almost-simple case is the only obstruction. The paper also gives a useful byproduct about independent primes in the soluble graph. However, the central almost-simple argument is not presented in a fully verifiable form: the authors explicitly state that detailed checking work is omitted, and many rows of Tables 1–14 are justified only by phrases such as 'handled exactly as a1)' or 'easily checked'. Because the Main Theorem is reduced exactly to Theorem 1, the correctness of every such row is load-bearing. The result is plausible and the structure is coherent, but as printed the proof is not independently checkable without redoing the classification data. In addition, one step in the Section 3 reduction, the assertion 'B ∩ M = 1' in step (vi), is not justified in the text.","major_comments":[{"comment":"The proof of Theorem 1 is the load-bearing part of the paper: Section 3(iv) reduces the Main Theorem to it. Yet the paragraph before Lemma 2 states that 'usually detailed checking work and easy calculations are omitted', and many entries in Tables 1–14 are approved by 'handled exactly as a1)', 'easily checked', or by reference to [13]. Two concrete examples: in §2.2.2(m), for N=U4(3), the proof that N_N(⟨y⟩) is the only maximal soluble subgroup of N containing an element of order 7 is asserted but not derived, and this uniqueness is essential for the contradiction; in §2.2.3(g), for N=PSp6(2), the claim that the only maximal soluble subgroups of N whose order is divisible by 15 are the normalizers of elements of order 5 is also asserted without proof. Since the minimal-counterexample reduction forces the almost-simple configuration, a single incorrect independence claim or a single missed factorization row would invalidate the Main Theorem. The manuscript should include the full verification for every row, or provide a reproducible machine-checkable supplement (for example explicit GAP or Magma checks against the ATLAS and the tables of [10] and [34]).","section":"Section 2 (Strategies; Tables 1–14)"},{"comment":"The assertion 'B ∩ M = 1' in the first paragraph of step (vi) is not justified in the text. From step (v) one only knows that there is some h with V_i,A=V_i,G and V_i,B=1 for i≤h, and V_i,B=V_i,G and V_i,A=1 for i>h. If h<k, elements of B∩M with nontrivial coordinates in positions h+1,...,k are not excluded, so B∩M=1 does not follow as written. To make the step valid one must first prove that h is either 0 or k: because G=AN and N acts trivially on the components, A acts transitively on {L_1,...,L_k}, and a full projection in one coordinate for A forces full projections in every coordinate; similarly for B. After swapping A and B if necessary, one may therefore assume h=k, and then B∩M=1 follows from V_i,B=1 for all i. As printed, this is a gap in a load-bearing step, although it is repairable by adding this argument.","section":"Section 3, step (vi)"}],"minor_comments":[{"comment":"The word 'explicitely' is misspelled twice; it should be 'explicitly'.","section":"Section 2.2 (introductory paragraphs)"},{"comment":"There is a typo 'subgoup' for 'subgroup'.","section":"Section 2.1 (Proposition 1)"},{"comment":"The sentence '7 divides |A ∩ N | and 3 4 divides |B ∩ N |' should read '3^4 divides |B ∩ N |'; the superscript appears to be missing.","section":"Section 2.2.2(m)"},{"comment":"The phrase 'whose order is divisible by 36, but not by 7' should read 'divisible by 3^6' rather than by the integer 36; as printed it is ambiguous.","section":"Section 2.2.3(b3)"},{"comment":"The notation 'V_{1,G} × ··· × V_{k,G}' is used as though it were a subgroup of M; a sentence explaining that this denotes the direct product in M = M_1 × ··· × M_k would improve clarity.","section":"Section 3, step (vi)"}],"recommendation":"major_revision","confidential_remarks":"The paper fits the scope of a group theory journal and the main theorem is attractive. The main barrier is verification of the almost-simple case analysis: the authors themselves state that detailed checks are omitted, and a referee cannot certify every row of Tables 1–14 without redoing substantial classification work. I would encourage the editor to request either a complete written verification or a computational artifact before acceptance. The Section 3 reduction is mostly convincing, but step (vi) needs the missing transitivity argument described in my report. There is no apparent circularity: the proof relies on independent classifications and prior radical-membership criteria, and the self-citations in [2], [4], [17] do not by themselves create a logical dependence on the present result."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Peter, if you work on products of finite groups, this is the natural next step after Carocca: G = AB is S-connected iff [A,B] lies in the soluble radical. That equivalence is genuinely new, and it pulls Thompson's theorem, Carocca's result, and the GKPS radical criterion into one statement. The byproduct, Theorem 2 on independent primes in almost simple groups, is also a real contribution to the prime-graph literature.\n\nWhere does this stand? I read Section 3 carefully. The minimal-counterexample reduction is coherent: unique non-abelian minimal normal subgroup, CG(N) = 1, N ≤ [A,B], then the wreath-product setup in (iv)-(vii). I did not find a gap. The alternating groups are worked in detail and are convincing, and the model linear-groups case a1) is shown honestly enough to follow the method.\n\nThe problem is Section 2. The authors say it themselves: \"usually detailed checking work and easy calculations are omitted.\" Everything reduces to Theorem 1 for almost simple groups, and Theorem 1 is a table-by-table run through LPS [34]. Most rows are justified by \"handled exactly as a1)\" or \"easily checked.\" Some rows are genuinely non-trivial: the U4(3) case rests on the claim that the normalizer of an order-7 element is the only maximal soluble subgroup containing it, and the PSp6(2) case on a uniqueness claim for maximal soluble subgroups of order divisible by 15. These are ATLAS checks, probably true, but not derived in the text. A single bad row would sink the main theorem, so as printed the proof is conditional: someone needs to re-run the rows.\n\nI would not call this a disqualifying problem. The authors are the people who know this classification material, the expositional standard is normal for this area, and they flag their own omissions rather than papering over them. If I needed the theorem I would cite it only after spot-checking the tables, but I would not bet against it. The citation pattern is fine: [18-21] and the forthcoming [17] are their own prior program, and the load-bearing outside references (LPS, ATLAS, Hering-Liebeck-Saxl, Abe-Iiyori, Amberg-Carocca-Kazarin) are appropriate.\n\nBottom line: send it to a serious referee. The right verdict is conditional acceptance with a request to supply the missing checks, or at least to prove the independence claims for the exceptional rows. The result deserves to be in the literature; it just needs to be checkable.","headline":"A genuine unification with a sound Section 3 reduction, but the almost-simple case analysis is an admitted outline: the theorem deserves publication, conditional on the tables being checked.","tokens_in":25271,"tokens_out":7968,"would_cite":true,"duration_ms":78646,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["20D40","20D10"],"pacs":[],"model":"deepseek-v4-flash","headline":"For a finite group written as a product of two subgroups, pairwise soluble generation is equivalent to the commutator of the factors lying in the largest soluble normal subgroup.","keywords":["Solubility","Products of subgroups","Two-generated subgroups","S-connection","Almost simple groups","Independent primes","Soluble radical","Finite groups"],"falsifier":"Take the exceptional almost simple types exhibited in the paper ($P Sp_6(2)$ and $U_4(3)$ with an outer automorphism of order $2$), construct their stated factorizations $G=AB$, and compute whether some $p$-element of $A\\cap N$ and $q$-element of $B\\cap N$ with $p\\neq q$ generate a non-soluble subgroup while condition (2) holds; if such a pair exists, the main theorem is false, and if the claimed independent primes are found, the theorem survives these hardest cases.","tokens_in":24320,"feed_emoji":"🧩","tokens_out":10842,"duration_ms":100245,"temperature":0.7,"pith_summary":"The paper establishes an exact structural counterpart of the classical two-generated criterion for solubility in the setting of factorized finite groups. It proves that for a finite group $G=AB$ with subgroups $A$ and $B$, the pairwise condition that $\\langle a,b\\rangle$ is soluble for every $a\\in A$ and $b\\in B$ holds precisely when the commutator $[A,B]$ lies in the soluble radical $G^S$, the largest soluble normal subgroup of $G$. The same equivalence already follows from a much weaker local test: it suffices to check pairs where $a$ is a $p$-element, $b$ is a $q$-element, and $p\\neq q$. This gives a local-global bridge of the kind that matters in finite group theory, and it implies, among other things, that a product of two soluble subgroups satisfying the pairwise condition is itself soluble.","feed_headline":"One commutator condition captures pairwise solubility in products","feed_subtitle":"In a finite group G = AB, testing elements of distinct prime order forces the commutator into the soluble radical.","key_machinery":"The central object is the soluble graph $\\Gamma_{\\mathrm{sol}}(N)$ of a non-abelian simple group $N$: its vertices are the primes dividing $|N|$, and two primes are adjacent exactly when $N$ contains a soluble subgroup whose order is divisible by their product. The proof's workhorse is the notion of an independent pair of primes, i.e. a non-edge of this graph. For every candidate factorization of an almost simple group, the paper uses an order comparison to force one specified prime into $|A\\cap N|$ and another into $|B\\cap N|$, and then uses maximal-subgroup structure, primitive prime divisors of $p^k-1$, and published subgroup data to show the two primes are independent; such a pair would contradict condition (2). The final contradiction for products with non-simple socle is obtained by projecting the factorization onto the simple direct factors and applying minimality.","core_discovery":"The main theorem states that the following three assertions about a finite group $G=AB$ are equivalent: (1) $A$ and $B$ are $\\mathcal S$-connected, meaning $\\langle a,b\\rangle$ is soluble for all $a\\in A$, $b\\in B$; (2) for every pair of distinct primes $p,q$, the subgroup $\\langle a,b\\rangle$ is soluble whenever $a\\in A$ is a $p$-element and $b\\in B$ is a $q$-element; and (3) $[A,B]\\leq G^S$, the soluble radical of $G$. The proof shows that a minimal counterexample to (2) implies (3) would have to be almost simple, with a non-abelian simple socle $N$ not contained in either factor, and then eliminates every possible factorization of such a group. The elimination is carried out by exhibiting, for each factorization, two primes that are independent with respect to $N$, one dividing $|A\\cap N|$ and the other dividing $|B\\cap N|$, where independence means that $N$ has no soluble subgroup of order divisible by their product. A separate theorem records that such independent primes occur for all almost simple factorizations except two explicit isomorphism types, with further small exceptions when independence is required with respect to the full automorphism group.","pith_inferences":["Because condition (2) only compares elements of different prime order, a finite list of prime-by-prime checks would decide $\\mathcal S$-connection for any concrete finite group, so the theorem gives a route to computer verification.","The same local-to-global shape suggests a family of testable conjectures: for other group classes with a well-behaved radical, connection of a product may be equivalent to $[A,B]$ lying in the corresponding radical, though the simple-group classification would need to be replaced by class-specific data.","The independent-primes theorem can be read as a reusable statement: for any factorization of an almost simple group outside the two exceptional types, a non-solubility certificate of the required kind is guaranteed to exist, which may simplify future arguments about products of almost simple groups.","A natural next step, suggested by the proof's own exceptional cases, would be a complete determination for alternating groups of which factorizations admit independent primes; the paper only gives partial results there."],"forward_implications":["If $A$ and $B$ are $\\mathcal S$-connected in $G=AB$, then $A^S=A\\cap G^S$ and $B^S=B\\cap G^S$; in particular, if $A$ and $B$ are soluble, then $G$ is soluble.","In an almost simple group, no nontrivial factorization can satisfy the distinct-prime condition: one of the factors must be trivial.","The theorem implies the known characterization of the soluble radical by pairwise solubility: an element $x$ lies in $G^S$ exactly when $\\langle x,y\\rangle$ is soluble for every $y\\in G$.","For deciding $\\mathcal S$-connection, only pairs of elements of distinct prime order need to be tested, so the condition is finite and checkable from Sylow-like data rather than from all pairs.","For almost all almost simple factorizations, the proof produces independent primes in the soluble graph, with only finitely many exceptions explicitly listed."],"supporting_citations":[{"why":"Supplies the complete classification of maximal factorizations of almost simple groups used to enumerate all candidate counterexamples.","marker":"[34]"},{"why":"Supplies the complete list of factorizations of the exceptional groups of Lie type.","marker":"[29]"},{"why":"Provides the lemmas on maximal soluble subgroups that establish independence of the chosen primes.","marker":"[2]"},{"why":"Provides subgroup structure data for small and exceptional groups used to check independence of primes in the remaining cases.","marker":"[13]"},{"why":"Provides maximal subgroup information for low-dimensional classical groups used in divisibility and independence checks.","marker":"[10]"},{"why":"States the original two-generated criterion for solubility that the paper extends to factorized groups.","marker":"[38]"},{"why":"Gives the characterization of the soluble radical by pairwise solubility used in the minimal counterexample step and recovered as a consequence.","marker":"[26]"},{"why":"Provides a solvability criterion used to conclude that the quotient G/N of a minimal counterexample is soluble.","marker":"[16]"},{"why":"Provides criteria for membership in the soluble radical via p-elements used in the non-simple socle step.","marker":"[25]"}],"fun_headline_variants":["Pairwise solubility in products reduces to commutator condition","Commutator criterion for solubility in finite products","Thompson's theorem extends to factorized groups","Prime-order pairs force commutator into soluble radical"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing assumption is that the list of possible ways an almost simple group can be written as a product of two subgroups is complete and correctly checked, including the exceptional finite cases; the paper says much of that detailed checking is omitted.","fun_headline_variants_meta":{"raw":{"variants":["Pairwise solubility in products reduces to commutator condition","Commutator criterion for solubility in finite products","Thompson's theorem extends to factorized groups","Prime-order pairs force commutator into soluble radical"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.001154,"raw_usage":{"total_tokens":4817,"prompt_tokens":1018,"completion_tokens":3799,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":634,"completion_tokens_details":{"reasoning_tokens":3740}},"tokens_in":634,"tokens_out":3799,"duration_ms":29678,"temperature":1.0,"reasoning_tokens":3740,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T14:16:51.473530+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Take the exceptional almost simple types exhibited in the paper ($P Sp_6(2)$ and $U_4(3)$ with an outer automorphism of order $2$), construct their stated factorizations $G=AB$, and compute whether some $p$-element of $A\\cap N$ and $q$-element of $B\\cap N$ with $p\\neq q$ generate a non-soluble subgroup while condition (2) holds; if such a pair exists, the main theorem is false, and if the claimed independent primes are found, the theorem survives these hardest cases.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplies the complete classification of maximal factorizations of almost simple groups used to enumerate all candidate counterexamples."},{"cited_title":"Hering, M","cited_arxiv_id":null,"evidence_quote":"Supplies the complete list of factorizations of the exceptional groups of Lie type."},{"cited_title":"Amberg, A","cited_arxiv_id":null,"evidence_quote":"Provides the lemmas on maximal soluble subgroups that establish independence of the chosen primes."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Provides subgroup structure data for small and exceptional groups used to check independence of primes in the remaining cases."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Provides maximal subgroup information for low-dimensional classical groups used in divisibility and independence checks."},{"cited_title":"Thompson, Nonsolvable finite groups all of whose loca l subgroups are solvable","cited_arxiv_id":null,"evidence_quote":"States the original two-generated criterion for solubility that the paper extends to factorized groups."},{"cited_title":"Guralnick, B","cited_arxiv_id":null,"evidence_quote":"Gives the characterization of the soluble radical by pairwise solubility used in the minimal counterexample step and recovered as a consequence."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Provides a solvability criterion used to conclude that the quotient G/N of a minimal counterexample is soluble."},{"cited_title":"Guest, D","cited_arxiv_id":null,"evidence_quote":"Provides criteria for membership in the soluble radical via p-elements used in the non-simple socle step."}],"review_version":1}